Cumulative Benefits of Extended-Release Memantine (28 mg, Once Daily) Across Clinical Domains in Patients With Moderate to Severe Alzheimer's Disease: An Area Under the Curve Analysis

Abstract

The efficacy and safety of the extended-release formulation of memantine (MemER; 28 mg, once daily) in patients with moderate to severe Alzheimer's disease (AD) have been demonstrated in a 24-week, randomized, double-blinded, placebo-controlled trial of 677 patients (placebo, n=335; MemER, n=342) concurrently receiving background therapy with a cholinesterase inhibitor (ChEI) (Grossberg et al, 2013). The protocol-specified analyses in Grossberg et al. focused on changes from baseline (Week 0) to endpoint (Week 24), and revealed significant benefits of adding MemER to the ongoing ChEI therapy in the domains of cognition (Severe Impairment Battery [SIB]), global clinical status (Clinician's Impression of Change Plus Caregiver Input [CIBIC-Plus]), behavior (Neuropsychiatric Inventory [NPI]), and semantic processing ability (Verbal Fluency Test [VFT]); no significant benefits were observed in the domain of daily function (the 19-item Alzheimer Disease Cooperative Study - Activities of Daily Living scale [ADCS-ADL19]). Instead of limiting analysis of treatment benefits to score changes between baseline (Week 0) and the Week-24 endpoint, which provides little information about variation among trial participants in their cumulative responses over time, this study aimed to explore treatment effects during the entire course of the trial. For each patient (observed cases; N=540), the area under the curve (AUC) for changes from baseline on the SIB, CIBIC-Plus, NPI, VFT, and ADCS-ADL19 was calculated for all available time intervals (Weeks 0-24, 4-24, etc), using the trapezoidal rule. For each treatment interval, patient-level data were combined and treatment groups were compared by means of an analysis of covariance (ANCOVA) model with treatment group, center, and baseline value in the model (α=0.05, two-sided). Over the entire study period (Weeks 0-24), the group treated with add-on MemER (MemER-ChEI) displayed AUCs for the SIB, CIBIC-Plus, and NPI that reflected mean improvements of 88% (71.2 ± 11.1 vs 37.9 ± 10.8, P=0.014), 133% (-5.3 ± 1.1 vs -2.3 ± 1.0, P=0.019), and 109% (-91.1 ± 11.9 vs -42.9 ± 11.5, P<0.001), respectively, over the AUC improvements observed in the placebo-ChEI group. (For CIBIC-Plus and NPI, negative score changes indicate improvement.) For the VFT, an AUC indicating cumulative improvement in the MemER-ChEI group was significantly different than the AUC for the placebo-ChEI group that indicated cumulative worsening (P=0.014); the cumulative worsening in the placebo-ChEI group was 139% greater in magnitude than the improvement observed in the MemER-ChEI group (2.5 ± 2.9 vs -6.0 ± 2.8). For the ADCS-ADL19, the mean AUC improvement in the MemER-ChEI group was 117% greater in magnitude than the AUC worsening observed in the placebo-ChEI group, but the difference was not statistically significant (3.9 ± 7.4 vs -1.8 ± 7.2, P=0.528). Other time intervals yielded similar results, i.e., significant AUC improvements for the MemER-ChEI group for the SIB, CIBIC-Plus, NPI, and VFT. In this post-hoc AUC analysis of a randomized placebo-controlled trial, cumulative benefits were observed across multiple clinical domains for the treatment of patients with moderate to severe AD with memantine ER added to background ChEI therapy. The AUC approach provides a more complete representation of longitudinal therapeutic effect than the change score analyses often used in clinical trials: it is more ecologically valid, robust, and dynamic; does not rely on simple two-point change scores; and is not limited by an assumption of linear changes with time in making inferences and interpreting results. The significant cumulative improvements in cognition, global clinical status, behavior, and semantic processing abilities provide support to the view that add-on therapy with memantine ER yielded meaningful therapeutic benefits across the 24-week trial period.